The Total Solar Eclipse of 1900. 81 
increasing the speed, until at the end (about 3 secs. after totality was over) the 
handle was turning at about the rate of one rotation per second. At this instant 
the increasing sunlight was so strong that, although the handle had been turned 
only ten times, and thus only ten exposures given, fearing lest some of the light 
might be reflected in the interior of the carriers, and thus fog the plates already 
exposed, I gave Sir Howard Grubb the signal to cap, just as the latter was about 
to do so on his own responsibility, as it had been arranged that he should do if it 
appeared to him that the exposures were being too much prolonged. We should 
thus, in any case, have lost the last two plates of this series; but, in addition, the 
oversight which had occurred in leaving one of the slides undrawn robbed us of 
half of the remaining ten, so that we have only five photographs relating to the 
latter phase of the eclipse. 
The twelve plates taken near second contact give an uninterrupted record 
of the changes in the chromospheric spectrum during the 17 or 18 seconds over 
which they extend. 
The instrument was focussed by means of a collimator, with a narrow slit in 
its focal plane. The collimator itself was focussed for parallel rays by means of 
an eye-piece behind the slit, placed so that the latter was exactly in its focus. 
The slit was then opened, and the collimator focussed as an ordinary telescope on 
a distant mountain peak. 
In focussing the camera, the Fraunhofer lines in the green and blue were made 
as sharp as possible, and then an allowance made for the difference in focus 
between rays corresponding to this part of the spectrum and those in the neigh- 
bourhood of H and K. Unfortunately this allowance was not hit off with perfect 
success, and in consequence the definition of the lines is not so sharp as might be 
desired, especially in the lower end of the spectrum. Towards the violet end the 
definition is better, but the sensitiveness of the plates (Sandell’s triple-coated) 
seems to fall off rapidly for wave-lengths beyond \3650. 
Several of the plates are disfigured by the effect of a slight jar which took 
place at one part of the movement, just as the light was being cut off one plate 
and turned on to another on the other side of the camera, and in nearly every case 
the brighter lines are accompanied by a straight wing, which seems to have been 
due to some irregular reflection from the moving mirrors. Neither of these defects, 
however, although they spoil the appearance of the plates very much, can be said 
to detract seriously from their scientific value. 
In the first two of these plates the F, H, and K lines of the chromosphere 
are exhibited boldly, H and K extending through an arc of nearly 180°, while F 
covers about 130°. The hydrogen lines, H, and H;, are the next in order of 
brightness, and extend through about 90°. Somewhat fainter than these are the 
H, and H, of hydrogen, the strontium line at 14078, and a line at 44023. Then 
TRANS, ROY. DUB, SOC., N.S., VOL. VIII., PART Y, 12 
